Beyond the Bounds of

Tarja Susi ([email protected]) Department of Computer Science, University of Skövde, Box 408 541 28 Skövde, Sweden

Jessica Lindblom ([email protected]) Department of Computer Science, University of Skövde, Box 408 541 28 Skövde, Sweden

Tom Ziemke ([email protected]) Department of Computer Science, University of Skövde, Box 408 541 28 Skövde, Sweden

Abstract stages of historical development humans went beyond the limits of the psychological functions given to them One of the questions that frequently come up in by nature and proceeded to a new culturally-elaborated discussions of situated, embodied and distributed organization of their behavior” (ibid. p. 39). cognition is where to draw the boundary between The question of the bounds of cognition received cognisers and their environment. Adams and Aizawa (2001) have recently formulated a critique of what they somewhat less attention during the first decades of consider a “radical view of tool use”, i.e., the view of cognitive science, which predominantly equated tools as part of the cognitive system. We analyse their cognition with internal computational processes critique and show that much of what they consider implemented by the brain and paid relatively little ‘radical’ turns out to be compatible with what they attention to the interaction of agents and environment. consider ‘common sense’. Hence, we argue that much of The question is currently going through a certain the debate boils down to a disagreement over different revival triggered by increasing interest in theories of uses of the term ‘cognitive’, whereas there is growing situated cognition (e.g., Clancey, 1997; Suchman, agreement about the central role that agent-environment 1987), embodied cognition (e.g., Clark, 1997; Varela, interaction in general, and tool use in particular, play in cognitive processes. We therefore suggest to drop the Thompson & Rosch, 1991) or distributed cognition ‘bounds of cognition’ debate, and conclude by raising (e.g., Hutchins, 1995), all of which emphasise the close what we consider more important questions in the study coupling between agent and environment and its central of cognitive tool use. role in cognitive processes. This shift in what is considered the appropriate unit Introduction of analysis in the study of cognition has noticeably also The question exactly where to draw the boundary led to a corresponding shift in the use of the term between a cognitive system and its environment is as ‘cognitive’. While the term traditionally has been used old as the study of mind itself. Polanyi (1964) and mostly for internal processing, in the 1990s it started to Bateson (1972) illustrated the question with the now appear in expressions like “cognitive tools” or classical example of a blind man using a stick, and “cognitive artifacts” (Norman, 1991, 1993). asked what the bounds of the blind man’s system are. Furthermore, several authors have started to More specifically, does it or does it not include the characterise the whole of humans and the technical stick? Another classical example, that of the knot in the tools they use, e.g., a pilot interacting with the handkerchief, comes from Vygotsky (1978), who instruments in her cockpit, or even a group of humans argued that the knot serves as a reminder that changes interacting with each other and the instruments on the one the psychological structure of the memory process, and bridge of a ship, as ‘cognitive system’ (Hutchins, it extends the operation of memory “beyond the 1995) or as a “joint cognitive system” (Hollnagel, in biological dimensions of the human nervous system” press). (ibid., p. 39). Vygotsky emphasised in particular the Others consider this a “radical view of tool use” role of cultural artefacts in both evolutionary and (Adams & Aizawa, 2001), which blurs the distinction individual development, elaborating that in “[t]he use of between cognitive agents and the non-cognitive tools notched sticks and knots, the beginnings of writing and they use (cf., for instance, Nardi, 1996; Neuman & simple memory aids all demonstrate that even at early Bekerman, 2000). Perhaps most notable among these critics are Adams and Aizawa (2001) who formulated a

1134 detailed critique of several theories they consider guilty manipulation so as to reduce the complex problem to a of going too far in blurring that distinction (Clark & sequence of simpler, pattern-completing steps that we Chalmers, 1998; Dennett, 1996; Donald, 1991; already command. On this model, then, it is the combination of our biological computational profile with Hutchins, 1995). They defend a ‘common sense’ view, the fundamentally different properties of a structured, which they refer to as ‘intracranialism’, considering symbolic, external resource that is a key source of our cognition as “restricted to the confines of our brains” peculiar brand of cognitive success. The external (Adams & Aizawa, 2001, p. 44). environment, actively structured by us, becomes a source We will here argue that Adams and Aizawa, as well of cognition –enhancing ‘wideware’– external items as other critics arguing along similar lines, might be (devices, media, notations) that scaffold and complement tilting at windmills, since most of the views that they (but usually do not replicate) biological modes of describe as ‘radical’ are in fact highly compatible with computation and processing, creating extended cognitive what they consider as ‘common sense’. systems whose computational profiles are quite different from those of the isolated brain” (Clark, 1999, p. 349, The Bounds of Cognition original emphasis). The general point of both these examples, that It is commonly agreed that humans use calculators, road cognitive processes can be complemented, augmented signs, notes, calendars, computers, pen and paper, and and transformed by environmental scaffolds, in even other people as external resources and as a way particular the use of tools, is relatively uncontroversial. around the limitations of their own cognitive capacities. Adams and Aizawa’s critique, however, is directed at In other words, as Clark (1997, p. 68) formulated it, we the idea of the environment as a “source of cognition” “call on external resources to perform specific and the characterisation of agent and environment as an 1 computational tasks” and thus depend on cultural “extended cognitive system”. According to them, artefacts to “augment and enhance biological cognition” “common sense has it that our cognitive faculties, (Clark, 1999, p. 350). restricted to the confines of our brains, can be aided in Consider, for example, the use of Scrabble tiles any manner of ways, by cleverly designed non- (Clark, 1997; Clark & Chalmers, 1998; Kirsh, 1997; cognitive tools” (Adams & Aizawa, 2001, p. 44, Kirsh & Maglio, 1994). As described by Clark (1997), emphases added). That means, they agree that the the tiles are physically ordered and re-ordered during coupling of internal cognitive processes with the play, thereby prompting our own on-line neural environment can augment those processes, but they resources. We manipulate the tiles externally and maintain that coupling “of some process with a broader thereby create a variety of fragmentary inputs (new environment … [does not] extend that process into the letter strings) capable of prompting the recall of whole broader environment” (ibid, p. 56). However, despite a words from the pattern-completing resource. It seems certain disagreement over the use of the term that our own biological resources do not easily provide ‘cognitive’, Adams and Aizawa’s position is highly for this kind of manipulations, which might therefore be compatible with, for example, that of Norman (1991, considered as a set of operational capacities that emerge 1993). Norman used the term ‘cognitive tools’ for tools from the interaction between brain and world. That that enhance human cognitive abilities, and never means, through the flexible use of environmental intended it to refer to tools literally having any of the resources we enhance or augment our own cognitive cognitive processes or abilities that humans are abilities – we use such resources as scaffolds. The endowed with. Considering processes as extending into Scrabble tiles, for instance, scaffold our thinking, and, the broader environment is not necessarily the same as thus in a very real sense it can be said that “the re- saying that the environment, or some part of it, actually arrangement of tiles on the tray is not part of action; it comes to have human-like cognitive processes or is part of thought” (Clark & Chalmers, 1998). capacities itself. Consider a second example, provided by Clark However, Clark and Chalmers (1998) who referred to (1999): their view of the extended mind as an “active “Most of us, armed with pen and paper, can … solve ” 2, argued that this is not just a matter of multiplication problems that would baffle our unaided terminology, but much more a matter of methodology: brains. In so doing we create external symbols “… in seeing cognition as extended one is not merely (numerical inscriptions) and use external storage and making a terminological decision; it makes a significant difference to the methodology of scientific investigation. In effect, explanatory methods that might once have been 1 Adams and Aizawa, as well as several of the authors they thought appropriate only for the analysis of ‘inner’ criticise, refer to cognitive and brain processes as processes are now being adapted for the study of the ‘computational’. This is, of course, not uncontroversial (cf., e.g., Clark, 1997), but the reader should not get distracted by this; the discussion in this paper is relatively independent of 2 To be distinguished from the ‘passive’ externalism of, e.g., whether ‘computation’ is the appropriate term in all cases. Putnam (1975) and Burge (1979).

1135 outer, and there is promise that our understanding will similar to traditional cognitive science, except that the become richer for it” (Clark & Chalmers, 1998). scope of the system has been widened to include a Concrete examples of extending the application of collaborating set of artifacts and people rather than the traditional cognitive scientific methods and terminology narrow ‘man-machine’ dyad of cognitive science … from ‘inner’ to ‘outer’ processes can be found in Treating each node in a system as an ‘agent’ … leads to a problematic view of cognition. We find in distributed Hutchins’ (1995) work on distributed cognition. cognition the somewhat illogical notion that artifacts are Hutchins is concerned with cognition at a ‘higher’ level, cognizing entities. Flor and Hutchins (1991) speak of such as team performance in ship navigation, i.e. inter- ‘the propagation of knowledge between different individual rather than intra-individual cognition. individuals and artifacts’. But an artefact cannot know Analysing ship navigation, Hutchins (1995) showed anything; it serves as a medium of knowledge for a how multiple embodied biological brains combine with human” (Nardi, 1996, pp. 86-87). tools (sextants, alidades, etc.), and media (maps, charts, It should be noted that while Adams and Aizawa as etc.) during performance. The artefacts allow the human well as Nardi might very well be right about the users “to do the tasks that need to be done while doing differences between human and machine ‘computation’ the kinds of things people are good at: recognizing or ‘information processing’, their criticisms are patterns, modeling simple dynamics of the world, and nevertheless misguided in the sense that none of the manipulating objects in the environment” (ibid., p. criticised authors actually denied those differences. As 155). the reader might have noticed in the above quotes, In this type of analysis, Hutchins, as he pointed out Clark (1999), for example, also referred to biological himself, deliberately applied “the principal metaphor of computation and external resources as “fundamentally cognitive science – cognition as computation – to the different”. Furthermore, he made it clear that he views operation of this system” (ibid., p. 49). Adams and the environment as a “source of cognition”, but only in Aizawa (2001), however, argue that in doing so the sense that it complements, rather than replicates, Hutchins “threatens to depart from common sense, biological computation and processing. Similarly, toward Dennett’s radical transcranial cognition” (ibid, Hutchins (1995) explicitly pointed out that he uses the p. 45), because “[i]f cognition is simply computation notion “cognition as computation” as a “metaphor” in over representational states, and if one’s tools, such as the description of distributed cognitive systems, and he paper and pencil, form or contain representations, then never denied the differences between intra-individual one has a case for the radical view that, in at least some and inter-individual cognitive processes. Hence, much cases of tool use, cognition extends beyond the of what Adams and Aizawa (2001) characterise as a boundary of the brain” (ibid, p. 46). In Adams and “radical view of tool use” turns out to be less radical Aizawa’s view, than it might seem at first. “… the kinds of computational processes we find The same applies to Adams and Aizawa’s critique of operating over external representations, such as marks on Donald (1991), whose view they consider “in many a piece of paper … will turn out to differ from the kinds respects ... the same as Dennett’s and Clark and of computational processes that we find operating over Chalmers’” (p. 45). Donald’s theory is concerned with representations in brains. Compare the intracranial computation of the product of 347 and 957 from the ‘exograms’, or external representations, and the way computation of this product with pencil and paper. We they have impacted, in the course of evolution, the may assume that there are computational processes at architecture of human cognition, allowing to off-load work in both cases, but that these computational biological memory. Donald claims that “[t]he exis tence processes are different. In particular, the internal of exograms eventually changed the role of biological processes are cognitive computational processes, where memory in several ways”. While the first two only some of the computational processes in the evolutionary transitions increased the load on biological transcranial cases are cognitive. In particular, it will be memory, “the final step in this tremendous cognitive only the internal portions of the transcranial computation expansion might have reduced the load on some aspects that turn out to be cognitive” (Adams & Aizawa, 2001, p. 59). of biological memory, by gradually shifting many They further argue that it is ”obvious” that brain storage tasks onto the newly developed ESS [external processes are ”causally distinct” from the processes symbol storage]” (Donald, 1991, p. 320, original involved in tool use, such as ”moving beads up and emphasis). down on rods in an abacus, or pressing buttons on an According to Adams and Aizawa (2001), Donald electronic calculator” (Adams & Aizawa, 2001, p. 44). implicitly refers to psychological laws of human A similar critique was formulated by Nardi (1996) memory and those will not generally hold for external who argued that the conceptual framework of memory storage. However, in Donald’s (1991) distributed cognition description the biological and the external are two quite “… views people and things as conceptually equivalent; different things: while engrams refer to single entries in people and artifacts are ‘agents’ in a system. This is the biological memory system, exograms refer to single

1136 entries in the ESS, and are considered external memory Beyond the Bounds records of ideas. Even though both engrams and The close coupling between the individual and the outer exograms are described in similar terms, exograms do world is realised through two major cultural mediators not become biological, not even ‘implicitly’. In fact, in human cognition, namely tools and language Donald notes that systems of exogram storage are much (Preston, 1998; Vygotsky, 1978). However, the role of more flexible than engrams and thus a symbolic artefacts and tools have mainly been left unattended information environment frees us from wholly while language, for instance, has received far more depending on biological memory. He therefore 3 attention (Preston, 1998; Wynn, 1991) . In language concludes that a “cognitive system containing exograms research there is “a sophisticated body of theory on how will have very different memory properties from a utterances are constructed. Nothing comparable exists purely biological system” (Donald, 1991, p. 315). for tool behaviour … [t]here is almost no concern with Obviously, this is much in line with Clark’s (1999) how tools are made and used and there are no well- aforementioned argument that extended cognitive developed theories of how sequences of tool-use are systems of biological brains and external resources constructed” (Wynn, 1993, p. 392). Others have also perform very differently than the ‘naked brain’ on its pointed out that development and tool use is largely an own. However, neither Donald nor Clark argue that overlooked issue in cognitive development (e.g., hooking the biological system to external resources Gauvain, 2001; Smitsman, 1997). However, artefacts transforms the external into biologically cognitive and tools have a similar role in cognitive processes as entities, but in fact both are careful to point out the that of language, “in particular they constitute the other fundamental differences. major form of cognitive mediation between individual Bateson (1972), in the example of the blind man and world” (Preston, 1998, p. 514). Yet, we do not fully using a stick, argued that questions concerning whether understand the relation between cognition and artefacts, a mental system is bounded by skin or skull, whether and there are several unanswered questions. For artefacts should be included or not, and so on, are in instance, how should we understand the concepts fact “nonsense questions”. Polanyi (1964) argued that “The way we use a hammer or a blind man uses his stick, ‘artefact’ and ‘tool’? While both terms can be grappled shows in fact that in both cases we shift outwards the with intuitively (or folk-psychologically), there do not points at which we make contact with the things that we seem to be any coherent definitions of them. A related observe as objects outside ourselves. While we rely on a question, as pointed out by Preston (1998), concerns tool or a probe, these are not handled as external objects what objects and behaviours should count as tools and … We pour ourselves out into them and assimilate them tool use, respectively. as parts of our own existence.” (Polanyi, 1964, p. 59) For further development of theories concerning Similarly, Bateson described the blind man’s stick as artefacts and cognition we should attend, for instance, “a pathway along which transforms of difference are research on tool making and tool use that has been being transmitted”, and “the way to delineate the conducted in the field of primatology (e.g., Boesch & system is to draw the limiting line in such a way that Boesch 1993; McGrew, 1992; Taylor Parker, Mitchell you do not cut any of these pathways in ways which & Lyn Miles, 1999). According to Tomasello (2000), leave things inexplicable” (p. 465). Subsequently, the human cognition is a particular form of primate boundaries of cognition, as well as the appropriate units cognition, since many structures of human cognition are of analysis in the study of cognition, depend on what identical with non-human primate cognition. Tomasello we want to explain. Human activities, for the most part, (2000) therefore argues that the study of non-human cannot be studied without considering things like the cognition can provide important information to artefacts we use. Neither can artefacts be studied in cognitive scientists . Research in ape-language has, for isolation since in itself a tool is ‘nothing’ (Ingold, instance, led to insights concerning the nature of 2000). Hence it is only natural for cognitive scientists language that might have been overlooked had we only interested in situated, embodied and distributed focused on language in human children (e.g., Savage- cognition to choose as their units of analysis cognitive Rumbaugh et al., 1998). Likewise, we might miss out agents in situ, i.e. embedded in their environments. aspects of tool use unless we take into consideration Whether or not such units of analysis should be referred findings in the field of primate tool use. The other way to as extended cognitive systems, and whether or not around, primatologists have attended findings in the tools involved should be referred to as cognitive cognitive science, e.g., by taking distributed cognition artefacts, is only a secondary question. What is as a framework for analysis of social interactions important in the notion of ‘cognitive’ artefacts or systems is the recognition of a close coupling between 3 In the field of AI, for example, there are thousands of papers the individual’s internal cognitive processes and the on AI models of language, but hardly any studies of AI outer world. systems’ tool use (i.e. the use of tools by AI systems, rather than the use of AI systems as tools).

1137 among non-human primates (Forster, 2002; Johnson, References 2001; Strum, Forster & Hutchins, 1997). Such cross- Adams, F. & Aizawa, K. (2001) The Bounds of fertilisation of areas (like cognitive science and Cognition. Philosophical , 14 (1), 43-64. primatology) might be advantageous for both fields. Bateson, G. (1972) Steps to an Ecology of Mind. When it comes to artefacts in the context of human Northwale, New Jersey: Jason Aronson Inc. activities, a lot of studies have focused on the individual Boesch, Ch. & Boesch, H. (1993) Diversity of Tool Use level and it is commonly recognised that and Tool-making in Wild Chimpanzees. In: A. “… the inclusion of a tool in the process of behavior … Berthelet & J. Chavaillon (Eds.) The Use of Tools by abolishes and makes unnecessary several natural Human and Non-human Primates. Oxford: Clarendon processes, whose work is accomplished by the tool; and Press. alters the course and individual features (the intensity, Burge, T. (1979) Individualism and the Mental. duration, sequence, etc.) of all the mental processes that Midwest Studies in enter into the composition of the instrumental act, , 4, 73-122. replacing some functions with others (i.e., it re-creates Clancey, W. J. (1997) Situated Cognition: On Human and reorganizes the whole structure of behavior just as a Knowledge and Computer Representations. New technical tool re-creates the whole structure of labor York: Cambridge University Press. operations)” (Vygotsky, 1981, pp. 139-140, in Cole & Clark, A. (1997) Being There: Putting Brain, Body, and Wertsch, 1996). World Together Again. Cambridge, Mass.: MIT However, we need to consider the role of artefacts Press. beyond the level of the individual - the question is, what Clark, A. (1999) An Embodied Cognitive Science? kind of replacements do artefacts cause at a multi- Trends in Cognitive Science, 3 (9), 345-351. individual level? We know that artefacts provide, e.g., Clark, A. & Chalmers, D. (1998) The Extended Mind. an external means for organising cooperative behaviour Analysis, 56, 10-23. (Hutchins, 1995; Susi & Ziemke, 2001), but how do Cole, M. & Wertsch, J. (1996) Beyond the Individual- artefacts effect cognition in shared activities? social Antinomy in Discussions of Piaget and Distributed cognition clearly has taken a step towards Vygotsky. Human Development, 39 (5), 250-256. explaining cognition at the multi-individual level. Dennett, D. (1996) Kinds of Minds. New York, NY: However, further work is needed in order to understand Basic Books. Origins of the Modern Mind: Three the role of artefacts in ‘higher’ level cognition. Donald, M. (1991) Stages in the Evolution of Culture and Cognition. Conclusions Cambridge, Mass.: Harvard University Press. Flor, N. V. & Hutchins, E. (1991) Analyzing Should the bounds of cognition be drawn so as to Distributed Cognition in Software Teams: A Case include the artefacts we use? From the classical Study of Team Programming During Perfective information processing view of cognition the answer is Software Maintenance. In: Empirical Studies of clearly ‘no’: cognition takes place solely within the Programmers: Fourth Workshop (pp. 36-64). boundaries of the brain. From a situated view of Forster, D. (2002) Consort Turnovers as Distributed cognition, however, quite a different picture emerges: Cognition in Olive Baboons: A Systems Approach to cognition is not purely what goes on inside the brain. Mind. In: M. Bekoff, C. Allen & G. M. Burghardt Rather, cognition emerges from the interaction between (Eds.) The Cognitive Animal: Empirical and brain, body and the environment (Clark, 1997), and in Theoretical Perspectives on Animal Cognition. order to understand cognition, the units of analysis in Cambridge, Mass.: MIT Press. cognitive studies need to be extended to include Gauvain, M. (2001) Cultural tools, social interaction Human external resources brought into our activities. In and the development of thinking. Development addition, we need to take notice of findings in other , 44, 126-143. Hendriks-Jansen, H. (1996) Catching Ourselves in the fields of research, such as primatology, to a greater Act: Situated Activity, Interactive Emergence, extent than presently done. As shown in this paper, the Evolution, and Human Thought. Cambridge, Mass.: inclusion of artefacts as part of ext ended cognitive The MIT Press. systems does not necessarily lead to illogical or absurd Hirose, N. (2002) An Ecological Approach to assumptions of artefacts coming to have human Embodiment and Cognition. Cognitive Systems cognitive processes or abilities, as some have argued. It Research, 3, 289 – 299. could therefore be concluded that debating where to Hollnagel, E. (in press) Cognition as Control: A draw the bounds of cognition to some extent simply is Pragmatic Approach to the Modelling of Joint ‘much ado about nothing’. Where the boundary is Cognitive Systems. IEEE Transactions on Systems, drawn is not the main issue – more importantly, we Man, and Cybernetics, to appear. need to attend the role of artefacts themselves in Hutchins, E. (1995) Cognition in the Wild. Cambridge, cognition. MA: MIT Press.

1138 Ingold, T. (2000) The Perception of the Environment: Comparative Perspectives. Cambridge: Cambridge Essays in Livelihood, Dwelling and Skill. London: University Press. Routledge. Tomasello, M. (1999) The Cultural Origin of Human Johnson, C. M. (2001) Distributed Primate Cognition: Cognition. Cambridge, MA: Harvard University A Review. Animal Cognition,3 (4), 167-183. Press. Kirsh, D. (1995) The Intelligent Use of Space. Artificial Tomasello, M. (2000) Primate Cognition: Introduction Intelligence, 73, 31-8. to the Issue. Cognitive Science, 24 (3), 351-361. Kirsh, D. (1996) Adapting the Environment Instead of Varela, F. J., Thompson, E. and Rosch, E. (1991) The Oneself. Adaptive Behavior, 4 (3/4), 415-452. Embodied Mind: Cognitive Science and Human Kirsh, D. & Maglio, P. (1994) On Distinguishing Experience. Cambridge, MA: MIT Press. Epistemic from Pragmatic Action. Cognitive Science, Vygotsky, L.S. (1977) The Development of Higher 18 (4), 513-549. Psychological Functions. Soviet Psychology, 16, 60- McGrew, W.C. (1992) Chimpanzee Material Culture: 73. Implications for Human Evolution. Cambridge: Vygotsky, L.S. (1978) Mind in Society: The University Press. Development of Higher Psychological Processes. Nardi, B. (1996) Context and : Activity Cambridge, Mass.: Harvard University Press. Theory and Human-Computer Interaction. London: Wynn, T. (1993) Layers of Thinking in Tool Behavior. MIT Press. In: K.R. Gibson & T. Ingold (Eds.) Tools, Language Neuman, Y. & Bekerman, Z. (2000) Where a Blind and Cognition in Human Evolution. Cambridge: Man Ends: Five Comments on Context, Artifacts and Cambridge University Press. the Boundaries of the Mind. Systems Research and Behavioral Science, 17, 315 – 319. Norman, D. (1991) Cognitive Artifacts. In: J.M. Carroll (Ed.) Designing Interaction: Psychology at the Human-Computer Interface (pp. 17-38). Cambridge: Cambridge University Press. Norman, D. (1993) Things That Make us Smart: Defending Human Attributes in the Age of the Machine. Cambridge, Mass.: Perseus Publishing. Polanyi, M. (1964) Personal knowledge: Towards a post-critical philosophy. New York: Harper Torchbooks. Preston, B. (1998) Cognition and Tool Use. Mind & Language, 13 (4), 513 – 547. Putnam, H. (1975) The Meaning of ‘Meaning’. In: K. Gunderson (Ed.) Language, Mind, and Knowledge. Minneapolis : University of Minnesota Press. Smitsman, A.W. (1997) The Development of Tool Use: Changing Boundaries Between Organism and Environment. In: C. Dent-Read & P. Zukow- Goldring (Eds.) Evolving Explanations of Development. Washington: American Psychological Association. Strum, S. C., Forster, D. & Hutchins, E. (1997) Why Machiavellian Intelligence may not be Machiavellian. In: A. Whiten & R. Byrne (Eds.) Machiavellian Intelligence II: Extensions and Evaluations (pp. 50- 85). Cambridge: Cambridge University Press. Suchman, L. A. (1987) Plans and Situated Actions: The Problem of Human-Machine Communication. New York: Cambridge University Press. Susi, T. & Ziemke, T. (2001) Social Cognition, Artefacts, and Stigmergy: A Comparative Analysis of Theoretical Frameworks for the Understanding of Artefact-mediated Collaborative Activity. Cognitive Systems Research, 2 (4), 273-290. Taylor Parker, S., Mitchell, R.W. & Lyn Miles, H. (1999) The Mentalities of Gorillas and Orangutans:

1139